4 Main Uses of Modeling in Engineering Design

The AnyBody Modeling System

John Rasmussen , in DHM and Posturography, 2019

2 The model repository

CAE models traditionally address artifacts such as cars, planes, and bridges, and they are developed bottom-up for each case. The human body is morphologically much more complex than most artifacts, so developing a human body model for each computation would be a major obstacle. On the other hand, despite the obvious differences between individuals, all humans are similarly composed in terms of body parts, joints, muscles, and so on, so it is feasible to provide a template model from which all other models can be produced by scaling, morphing, or other modifications.

Development of a valid template of the human anatomy is an open-ended task, partially because open questions on the composition and function of many elements of the musculoskeletal system remain. Furthermore, models for different purposes require different levels of detail, and there is probably no end to the level of detail to which the body can be modeled with continued effort. For instance, modeling the knee joint as a hinge may be a valid assumption for an overall posture or movement analysis, while it is probably insufficient for a detailed assessment of cruciate ligament loads.

It is vital that users have the opportunity to scrutinize models and make informed decisions as to their applicability for a given purpose. This opportunity can probably only be provided by open models. This principle is the foundation for the AnyScript Model Repository (Lund, Tørholm, & Jung, 2018). This repository is managed by AnyBody Technology to ensure the mutual structure and connectivity of body parts, but the models are stored in text files in an open repository (Lund et al., 2018) and are available for investigation by the research community and the public in general.

Models developed for a simulation of a particular situation are called "Applications" (Fig. 8.1) in AMS terminology. They are typically built from two primary elements: (1) a body model that is imported from the repository and possibly modified to represent an individual or a statistical variation over a population, and (2) an environment that the body is interacting with, for instance, a floor, force platforms, a chair, a bicycle, a medical device, or a piece of sports equipment. The environment is sometimes the result of an engineering design process and can be imported from a computed-aided design (CAD) system. Fig. 8.2 shows the composition of an AMS application.

Figure 8.2. Elements forming an AMS application. AMS, AnyBody Modeling System.

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B9780128167137000088

Design Guidelines

In Hybrid Microcircuit Technology Handbook (Second Edition), 1998

5.5 Engineering-Model Design Verification

The first engineering-model circuit is integrated with the test adapter and tested according to the requirements of the Functional Test Specification. Deviations between the test requirements and circuit performance are identified. The test results are analyzed to determine if the deviations are related to marginal circuit design, test adapter design, or unrealistic specification requirements. The system user is requested to evaluate the deviations between actual and specified performance with regard to required system performance and, when possible, to integrate the circuit into the next assembly level to verifjl system performance.

The engineering-model circuit performance should also be verified over the system operating temperature range. Temperature testing is particularly important for those applications requiring production testing at temperature. The test results are used to aid in establishing test limits at temperature.

The design verification is also used to perform any special tests such as design margin, limit testing, and noise testing.

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B9780815514237500122

Underground coal gasification research and development in the United States⁎

D.W. Camp , in Underground Coal Gasification and Combustion, 2018

4.6.3 Simplified engineering models

Simpler engineering models were developed that are easier to use by a competent UCG engineer and are useful for obtaining rough estimates, dependencies/sensitivities, and screening resources. These are typically lumped-parameter models with no spatial or temporal resolution.

LLNL developed EQSC (Upadhye, 1986) to calculate energy and material balances (by species) based on a simple multizone model of UCG, chemical equilibrium, and a set of required inputs. In addition to the coal analysis, inputs included the water influx and the fraction of this that enters the process before and after the water-gas shift equilibrium is set, the methane ratio (since methane is governed more by pyrolysis than equilibrium), the effective temperature for calculating water-gas-shift equilibrium (different than the process temperature), heat loss, and product exit temperature. EQSC was extended in recent years by LLNL to a spreadsheet-based model called UCG-MEEE (material, energy, and economics estimator) (Upadhye et al., 2013). Its core is the EQSC model for a single (but reproducible) UCG module. UCG-MEEE provides side calculations to help estimate some of the required input parameters. It envisions many UCG modules operating in parallel on industrial scale for a long project duration, and so, it requires and calculates industrially relevant engineering parameters. In addition to a full material (by species) and energy balance and large-scale flow and resource parameters, UCG-MEEE estimates a selling price for the product gas to achieve a desired rate of return. The economics are based on the GasTech (2007) study and standard scaling factors. UCG-MEEE's utility for determining parameter sensitivities and trade-offs was illustrated in Burton et al. (2012).

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B9780081003138000049

Pitting corrosion

Sankara Papavinasam , in Trends in Oil and Gas Corrosion Research and Technologies, 2017

28.4 Corrosion engineering models

Corrosion engineering models to predict internal pitting corrosion of oil and gas pipelines are based on nonclassical theory of pitting corrosion. According to these models the internal pitting corrosion of carbon steel in the oil and gas industry operating conditions occurs in four stages ( Fig. 28.1):

Figure 28.1. Stages of pitting corrosion according to the corrosion engineering model (Papavinasam model).

1.

The low- or no-corrosion stage when the internal surface of the pipeline is covered by hydrocarbons, i.e., oil-wet conditions,

2.

Formation of surface layers on the steel surface due to corrosion reactions once the surface is covered with water, i.e., water-wet conditions,

3.

Initiation of pits at localized regions on the steel surface when surface-layer breakdown occurs, and

4.

Pit propagation and eventual penetration of the pipe wall.

Although some corrosion engineering models address one or two stages of internal pitting corrosion of carbon steels, only one model [13–16] integrating all four stages is described in this section.

28.4.1 Stage 1

Oil and gas industry infrastructures do not suffer from corrosion if the surface is wetted and covered with oil or any hydrocarbon. But during operation, water may enter the system. Water enters into a pipeline from various sources:

•

A new pipeline, after construction, is hydrotested to ensure its integrity. If the pipe is not dried properly after hydrotesting, water may accumulate at low-lying regions.

•

In a production pipeline, water enters from the formation well. Initially the volume of formation water being transported in the pipeline is low, but as production continues the amount of water transported increases.

•

Water content of a gas production pipeline depends on the moisture content of the gas. Below dew point, the water condenses out from the gas and accumulates at low-lying regions.

•

Transmission pipelines in principle should not carry more than 0.5   vol% of water. Over years of operation, however, water accumulates in low-lying regions of the pipeline.

Whatever the source of water in pipelines, whenever it comes in contact with the surface of pipeline material (typically carbon steel) the surface becomes susceptible to corrosion. Hydrocarbons play an important role in determining whether the water will come in contact with the surface and in determining the corrosivity of the water.

Because of their nonionic nature, hydrocarbons cannot dissolve in ionic water. However, at low concentrations of water, hydrocarbons can form emulsion with the water. The type of emulsion and its stability depend on the type of hydrocarbon, the ionic content of the water, as well as the pressure, temperature, and flow rate.

There are two kinds of emulsions: oil-in-water (o/w) and water-in-oil (w/o). In a w/o emulsion, oil (hydrocarbon) is the continuous phase; therefore, w/o has low conductivity and is noncorrosive. In an o/w emulsion, water is the continuous phase. Therefore o/w has high conductivity and is corrosive. In an operating pipeline, initially the amount of water carried is lower and the amount of oil carried is higher, and the water content progressively increases. The percentage of water at which w/o converts to o/w is known as the emulsion inversion point.

The presence of free water, or of o/w emulsion, does not necessarily lead to corrosion. Under this condition, wettability of the hydrocarbon on the carbon steel determines corrosivity. Based on the wettability, oil can be classified into three categories [13]:

•

Oil-wet surface: The oil has a strong affinity to be in contact with carbon steel. Oil-wet surfaces physically isolate the pipe from the corrosive environment, and under such conditions, corrosion does not occur.

•

Water-wet surface: The oil does not have an affinity to be in contact with carbon steel; in fact the oil may not be in contact with the carbon steel at all, even when it is the only phase. A water-wet surface (in the presence of oil) is highly susceptible to corrosion.

•

Mixed-wet surface: The oil does not have any preference to be in contact with carbon steel. The oil may be in contact with the carbon steel surface as long as there is no competing phase present.

In the presence of free water or o/w emulsion, and on a water-wet surface, hydrocarbon can influence the incidence of corrosion in the water phase. If the hydrocarbon contains water-soluble corrosion inhibitors, it could either prevent (preventive hydrocarbon) or decrease (inhibitory hydrocarbon) corrosion. On the other hand, if the hydrocarbon contains water-soluble corrosive substances, it could increase the corrosion rate (corrosive hydrocarbon). If the hydrocarbon does not contain any water-soluble substances, or substances that could adsorb on steel, it will not have any influence on the corrosivity of brine (neutral hydrocarbon).

The factor responsible for one property may or may not influence other properties. For this reason, all three properties (emulsion, wettability, and corrosivity) are required, and based on them the influence of hydrocarbons can be predicted (Fig. 28.2).

Figure 28.2. Predicting influence of hydrocarbons on internal corrosion of pipelines.

Adopted from ASTM G205.

In stage 1, the PCR will be low, i.e., PCR 1 in Fig. 28.1, as long as the oil is wet and is protecting the surface.

28.4.2 Stage 2: surface layers

Once the surface becomes water-wet, corrosion takes place leading to formation of surface layers. Under the oil and gas operating conditions iron sulfide (FeS), iron carbonate (FeCO₃), iron oxides (FeO, Fe₂O₃, Fe₃O₄), biofilms, and any combination of them can form. Both composition of the surface layers and morphology of their formation are important. The composition and stability of the layers depend on various factors, including temperature, brine composition, oil composition, steel composition, pressure, velocity, microbes, corrosion inhibitors, and biocides [14].

In stage 2, the corrosion rate will be high at the beginning, i.e., PCR 2 in Fig. 28.1, and as more and more protective layers are formed, it will progressively decrease. When the surface layer formation reaches the steady state the PCR will be low, i.e., PCR 3 in Fig. 28.1.

28.4.3 Pit initiation

On a surface layer–protected surface, pits can be randomly initiated. In a study, experiments have been carried out in a high-temperature, high-pressure rotating cylinder electrode (HTHPRCE) apparatus under 40 conditions to cover the spectrum of conditions found in oil and gas production pipelines in Western Canada [15]. Each experiment was conducted over a period of 100   h. During the experiments, the electrodes were monitored using electrochemical noise technique. After the experiments, the surfaces of the samples were analyzed using a scanning electron microscope (SEM). Potentiodynamic polarization (PP) experiments were also conducted under the same 40 conditions.

Based on the electrochemical noise, SEM, and PP data, it was found that the probability of the initiation of pitting corrosion:

•

increased with increase in flow rate, temperature, and chloride ion concentration and

•

decreased with increase in oil wettability, H₂S partial pressure, CO₂ partial pressure, total pressure, bicarbonate concentration, and sulfate concentration.

The variation in the probability of initiation of pitting corrosion depends on the type of surface layers formed. Depending on the type of surface layer characteristics, the PCR can be high, i.e., PCR 4 in Fig. 28.1, may be higher than PCR 2.

28.4.4 Pit propagation

When the pits become sufficiently deep, they continue to grow until failure occurs; sometimes the growth rate will accelerate (autocatalytic process), i.e., PCR 5 in Fig. 28.1 can be higher than PCR 4. The depth of a corrosion pit depends on the pit growth rate and the timing of its initiation. Based on experiments conducted in six operating oil and gas production pipelines over a period of 4   years, internal pit growth rates under realistic operating conditions have been determined [16]. The study concluded the following:

•

Pit growths in both horizontal and vertical pipelines were similar when the compositions of surface layers were the same.

•

When compact layer of single species was formed, the surface was protected from pitting corrosion and that FeS-covered surfaces were less susceptible to pitting corrosion than FeCO₃-covered surfaces.

•

The pit growth rate increased when the surface layers of multiple species were formed. Sudden changes in the operating conditions appeared to facilitate formation of layers of multiple species, leading to higher pit growth rate.

•

In the absence of surface layers, the pit growth rate decreased but not eliminated, because even materials not adherent on to the surface might create uneven distribution of anodic and cathodic areas.

According to the corrosion engineering model described in this chapter, operational parameters required to predict PCR s are (1) production rate of oil, (2) production rate of water, (3) production rate of gas, and (4) production rate of solid, (5) total pressure, (6) partial pressure of CO ₂, (7) partial pressure of H₂S, (8) temperature, and (9) concentration of chloride ion, (10) concentration of sulfate ion, and (11) concentration of chloride ion. Each one of the operational parameters individually can alter the PCRs. The ultimate rate at which the pits will propagate depends on the combined effect of all of the operational parameters. Although the individual effect of each of the parameters can be predicted deterministically, determining the combined effect of these variables needs application of statistical principles. Several methods have been used to predict the probability of long-term PCRs based on short-term experiments. In all these approaches, one commonality is the acceptance that the driving force for the pitting corrosion is a "distributed parameter"; the different approaches vary on how "driving force" and "pattern of distribution" are treated.

It is assumed that each operational variable produces an individual pit growth rate (resulting in 11 different PCRs) and that pit growth rate as a result of the variables not considered (e.g., acetic acid effect) is the mean value of these 11 pit growth rates. The 12th parameter does not have any effect on the predicted PCR; it does, however, increase the uncertainty of the prediction. The actual pit growth rate taking place in the oil and gas pipeline is the "distributed function," which is the mean value of the 12 pit growth rates and the uncertainty is expressed as standard deviation. The resultant pit growth rate is the rate at which the pits will start to grow in the localized anodic region where the surface layers are removed. It should be noted that except for one operational variable (i.e., water), removal or control of other variables, does not prevent pit growth although it does change the pit growth rate. The absence of water prevents pitting corrosion regardless of the effect of other variables. In the model described in the chapter, the uncertainty is expressed as standard deviation.

Pits will not continue to grow at the same rate at which they start growing, for various reasons including partial reformation of the surface layers, local solution saturation, change of corrosion potential, and local increase of pH at these locations. As a result, the pit growth rate diminishes parabolically as a function of time.

Thus based on laboratory experiments conducted under pipeline operating conditions and assumptions made on the probabilistic nature of the pitting corrosion, the PCR of oil and gas pipelines can be calculated using Eq. (28.20):

(28.20) $\begin{array}{c}\text{Pitting}\text{Corrosion}\text{Rate}\left(\text{mpy}\right)=\{[\sum (-0.33\mathit{\theta }+55)+(0.51\mathit{W}+12.13)+(0.19{\mathit{W}}_{\text{ss}}+64)+\\ (50+25{\mathit{R}}_{\text{solid}})+(0.57T+20)+(-0.081{\mathit{P}}_{\text{total}}+88)+\left(-0.54{\mathit{P}}_{{\text{H}}_2\text{S}}+67\right)+\\ (-0.013{\mathit{C}}_{\text{sulfate}}+57)+\left(-0.63{\mathit{P}}_{{\text{CO}}_2}+74\right)+(-0.014{\mathit{C}}_{\text{bicarbonate}}+81)+\\ (0.0007{\mathit{C}}_{\text{chloride}}+9.2)+\left(\mathit{C}\mathrm{.}\mathit{R}{\mathrm{.}}_{\text{general}}\right)]/12\}\end{array}$

where θ is the contact angle of oil in a water environment; W is (water production rate/water   +   oil production rates   ×   100); W _ss is wall shear stress; R _solid is equal to 1 if the pipeline has solids or 0 if the pipeline does not have solids pipe; T is temperature in degrees celsius; P _total is total pressure in psi; ${\mathit{P}}_{{\text{H}}_2\text{S}}$ is partial pressure H₂S in psi; ${\mathit{P}}_{{\text{CO}}_2}$ is partial pressure of CO₂ in psi; C _sulfate is sulfate concentration in ppm; C _bicarbonate is bicarbonate concentration in ppm; C _chloride is chloride concentration in ppm; C.R._general is the average pit growth rates in oil, water, flow, solid, temperature, total pressure, partial pressure of H₂S, partial pressure of CO₂, and chloride; and t is a constant depending on the time.

In developing this model, the following assumptions have been made:

•

The type of carbon steel used will not effect of the PCR.

•

The volume and type of solid production is assumed to be irrelevant.

•

CO₂ and H₂S concentrations are not zero.

•

Unless the actual emulsion type is determined, the emulsion is oil-in-water.

Within the constraints of the assumptions made and the range of conditions in which the experiments were conducted, Eq. (28.20) can be used to predict the internal pitting corrosion of sour and sweet carbon steel pipelines.

It should also be pointed out that the elements that drive the PCRs of oil and gas pipelines are the statistical functions and hence the predicted PCR is probabilistic in nature. The incidence error in the prediction was determined using the standard deviations of Eq. (28.20). The driving forces could also be distributed other statistical tools [17,18].

While Eq. (28.20) provides the trend in the variation of PCR when only one variable is changed, under operating conditions of the oil and gas production, some or all variables coexist and coinfluence the localized PCRs. The ultimate rates at which the pits would propagate depend on interactions between any or all of the variables. The interactions can be independent, synergistic, or antagonistic.

•

Independent: Presence of other variables does not affect the localized pitting corrosion caused by a particular variable. The model described in this chapter, considers that the variables are independent to one another.

•

Synergistic: Presence of other variables mutually increases the localized pitting corrosion caused by each one of them. The net result is that the corrosion rate will be higher than the corrosion rate calculated from the average of corrosion rates due to individual variable. This is also known as autocatalytic effect.

•

Antagonistic: Presence of other variables mutually decreases the localized pitting corrosion caused by each one of them. The net result is that the corrosion rate will be lower than the corrosion rate calculated from the average of corrosion rates due to individual variable [19–21].

The predominant types of bacteria associated with microbiologically influenced corrosion (MIC) are sulfate-reducing bacteria, sulfur oxidizing bacteria, iron oxidizing/reducing bacteria, manganese-oxidizing bacteria, acid-producing bacteria, and slime formers. These organisms coexist within a biofilm matrix on metal surfaces, functioning as a consortium, in a complex and coordinated manner. The various mechanisms of MIC reflect the variety of physiological activities carried out by these different types of microorganisms when they coexist in biofilms. Despite decades of study on MIC, how many species of microorganisms contribute to corrosion are still not known with certainty, and researchers continue to report on the formation of biofilms by an ever-widening range of microbial species.

Frequently, an engineer may be required to make an appraisal of the MIC threat with limited information on the corrosion history and with little or no historic microbiological test data. In such cases, a preliminary analysis of the system operating conditions may be sufficient to exclude the MIC threat when these conditions are not compatible for the survival of sessile bacteria/biofilm. With this approach, a qualitative risk score model has been developed. Table 28.2 presents the score according to the MIC model. The PCR_MIC (PCR_MIC in mpy or PCR_MIC in mm/y depending on units used in Eq. 28.20) including the influence of MIC is then calculated using Eq. (28.21) [22–24]:

Table 28.2. Risk scores for microbiologically influenced corrosion (MIC)

Influence of parameter	Range of parameter	Unit of parameter	MIC risk score a	Remarks
Temperature	Less than −10	°C	0
	−10–15		1
	15–45		7–10
	45–70		7–4
	70–120		4–1
	Above 120		0
Pressure	Greater than 20	${\mathit{P}}_{{\text{CO}}_2}/{\mathit{P}}_{{\text{H}}_2\text{S}}$ b	10	Only if gas contains H2S greater than 10   mol/kmol (1%) by volume.
	Less than 20		2
Flow rate	Above 3	m/s	1
	2–3		2–12
	1–2		12–18
	0–1		18–20
pH	Less than 1		0
	1–4		5
	4–9		10
	9–14		1
	Above 14		0
Langelier saturation index (LSI)	Less than −6		10
	−6–−1		10–5	MIC tendency decreases as the LSI value increases in the negative direction because the tendency of non-MIC increases.
	−1–1		0
	1–8		1–8	MIC tendency increases as the LSI value increases as more scales are formed.
	Greater than 8		8
Total suspended solids (TSS)	Present		10	If the flow rate is between 0 and 3   m/s.
	Present		0	If the flow rate is above 3   m/s.
	Absent		0
Total dissolved solids	Less than 15,000	ppm	1
	15,000–150,000		1–10
	Greater than 150,000		10
Redox potential (Eh)	Less than −15	mV	1
	−15 to +150		1–10
	Greater than 150		10
Sulfur content	Present		10
	Absent		1

a

Summation of all MIC factors. The maximum value is 100.

b

Partial pressure ratio.

(28.21) ${\text{PCR}}_{\text{MIC}}=\text{PCR}\times \left(\frac{\text{MIC}\_\text{Risk}\_\text{Score}}{50}\right)$

where PCR is the corrosion rate due to non-MIC activities (as calculated using Eq. 28.20) and MIC_Risk_Score is the MIC factor, calculated using Table 28.2.

When applying internal localized pitting corrosion model to a real operating environment, the dynamic aspects of the operating conditions must be taken into account. There are several variables to consider to account for the dynamic nature of field operation. However, four important variables are flow velocity, flow regime, operating boundaries, and time. The following sections describe how the four variables are treated in this model.

Depending on the pressure, temperature, pipe diameter, and elevation profile, the flow velocity varies. Flow velocity determines where water accumulates. Detailed procedures to determine flow velocities and to calculate locations where water may accumulate are available elsewhere [1].

Flow patterns of multiphase (transporting simultaneously oil, water, gas, and solid) pipelines are commonly known as flow regimes. The flow regimes depend on the diameter of the pipe, orientation of the pipe, the flow rates in the pipe, and fluid properties. Description of types of flow regimes and their characteristics are beyond the scope of this chapter but are available elsewhere [1]. Based on evaluation of several field data, general guidelines on the influence of flow regimes on corrosion have been established (Table 28.3). PCR_MIC determined by Eq. (28.21) is corrected with factors presented in Table 28.3 to account for flow regime to obtain PCR_mean (in mpy or in mm/y depending on unit used in Eq. 28.20).

Table 28.3. Variation of pitting corrosion rate as a function of flow regimes

Flow regime type	PCRMIC modification
Slug flow	No change
Plug flow	PCRMIC   ×   0.98
Bubble flow	PCRMIC   ×   0.96
Dispersed flow	PCRMIC   ×   0.94
Oscillatory flow	PCRMIC   ×   0.92
Annular flow	PCRMIC   ×   0.90
Churn flow	PCRMIC   ×   0.88
Wave flow	PCRMIC   ×   0.86
Stratified flow	PCRMIC   ×   0.84

PCR, pitting corrosion rate; MIC, microbiologically influenced corrosion.

Pits will not continue to grow at a constant rate for various reasons. To account for this effect the average pitting corrosion rate, PCR_(average) (in mpy or mm/y depending on unit used in Eq. 28.21) from PCR_mean is calculated using Eq. (28.22):

(28.22) ${\text{PCR}}_{\left(\text{average}\right)}=\frac{\frac{{\text{PCR}}_{\text{mean}}}{1}+\frac{{\text{PCR}}_{\text{mean}}}{2}+\frac{{\text{PCR}}_{\text{mean}}}{3}+\cdots \cdots +\frac{{\text{PCR}}_{\text{mean}}}{t}}{T}$

where "T" is the total number of years between the start date and current date, t is the number of times the PCR_mean should be calculated. For example, the system that is operating for 5   years, the value of T is 5, and the PCR_mean calculation should be carried out five times, i.e., in year 1, 2, 3, 4, and 5 (t).

The PCR calculated using Eq. (28.22) is valid only if the production conditions are constant over the years. If the production conditions change for a particular year the value of "t" is set to unity for that year, and the "t" values for subsequent years increase as per Eq. (28.22). Table 28.4 provides the boundary conditions to determine if the production conditions change or not. For a system with PCR_mean of 10   mpy (0.25   mm/y) operating within the boundary (See Table 28.4), the PCR_average over 5   years will be about 4   mpy (0.1   mm/y). For a system with same PCR_mean of 10   mpy (0.25   mm/y) but with operating conditions changing beyond the boundary conditions (see Table 28.4) in the third year, the PCR_average over 5   years will be about 7   mpy (0.18   mm/y). For a system with same PCR_mean of 10   mpy (0.25   mm/y) but with operating conditions changing beyond the boundary conditions (see Table 28.4) every year, the PCR_average over 5   years will be 10   mpy (0.25   mm/y).

Table 28.4. Boundaries to determine if the production conditions change or not

Parameter	Boundaries
Temperature, °C	X   ≤   25
	25   <   X   ≤   50
	X   >   50
Pressure, psi (kPa)	X   ≤   100 (689)
	100 (689)   <   X   ≤   500 (3447)
	X   >   500 (3447)
H2S, psi (kPa)	X   ≤   2.5 (17.24)
	2.5 (17.24)   <   X   ≤   10 (69)
	10 (69)   <   X   ≤   50 (345)
	X   >   50 (345)
CO2, psi (kPa)	X   ≤   2.5 (17.24)
	2.5 (17.24)   <   X   ≤   10 (69)
	10 (69)   <   X   ≤   30 (207)
	30 (207)   <   X   ≤   100 (689)
	X   >   100 (689)
${{\text{SO}}_4}^{2-}$ (ppm)	X   ≤   750
	750   <   X   ≤   1000
	1000   <   X   ≤   1500
	1500   <   X   ≤   2500
	X   >   2500
${{\text{HCO}}_3}^{-}$ (ppm)	X   ≤   500
	500   <   X   ≤   1000
	1000   <   X   ≤   2000
	2000   <   X   ≤   4000
	X   >   4000
Cl− (ppm)	X   ≤   10,000
	10,000   <   X   ≤   20,000
	20,000   <   X   ≤   40,000
	40,000   <   X   ≤   60,000
	60,000   <   X   ≤   80,000
	80,000   <   X   ≤   100,000
	100,000   <   X   ≤   120,000
	X   >   120,000

The prediction from the internal localized pitting corrosion model has been validated using data collected from 27 fields. Table 28.5 presents a comparison of localized PCRs predicted by the model and field maximum pitting corrosion rates. The validation data presented in Table 28.5 are only values at the time the data were collected. Any changes in the operating conditions will influence the predicted PCR.

Table 28.5. Validation of localized pitting corrosion model

Field a	$\mathbf{p}{\text{H}}_2\text{S}$ , psi (kPa) b	$\mathbf{p}{\text{CO}}_2$ , psi (kPa)	Predicted PCR, mpy (mm/y) b	Field maximum PCR, mpy (mm/y) b	References
1	0.2 (1)	320 (2206)	20 (0.45)	18 (0.45)	[16]
2	2 (14)	4 (28)	13 (0.33)	2 (0.05)	[16]
3	2 (14)	4 (28)	16 (0.40)	2 (0.05)	[16]
4	2 (14)	4 (28)	31 (0.78)	27 (0.68)	[16]
5	0.2 (1)	2 (14)	16 (0.40)	11 (0.28)	[16]
6	0.2 (1)	0.1 (1)	16 (0.40)	17 (0.43)	[16]
7	60 (414)	6 (41)	47 (1.18)	48 (1.20)	[20,21]
8	0 (0)	21 (145)	0.7 (0.02)	0.1 (0.00)	[22]
9	0 (0)	20 (138)	1.2 (0.03)	0.1 (0.00)	[23]
10	0 (0)	21 (145)	0.5 (0.01)	0.1 (0.00)	[24]

a

Validation of the model for another 17 fields have previously been presented [1].

b

See original papers for other operating conditions and for pitting corrosion rate distributions.

The model presented in this chapter is validated only under certain conditions, and hence should only be used within the conditions. Table 28.6 provides the boundary conditions for this model.

Table 28.6. Boundary for using the model (applicable to carbon steel only)

Experimental parameter	Minimum condition	Maximum condition	Units
Solid	No solids present	Solids are present	n.a
Temperature	5	65	°C
Total pressure	0	750 (5171)	Psi (kPa)
H2S partial pressure	0	50 (345)	Psi (kPa)
CO2 partial pressure	0	100 (689)	Psi (kPa)
Bicarbonate concentration	0	4000	Ppm
Sulfate concentration	0	2500	Ppm
Chloride concentration	0	120000	Ppm
Flow	0	5	m/s

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B9780081011058000280

Advanced simulation techniques in vehicle noise and vibration refinement

N. Hampl , in Vehicle Noise and Vibration Refinement, 2010

8.4.2 Noise and vibration refinement deliverables for virtual series

Each component CAE model that is required for a full vehicle noise and vibration simulation model will first be checked and analysed by the responsible component area to ensure that the component will meet its targets. Typical noise-relevant component attributes are dynamic stiffness, component eigenfrequencies and mode shapes, noise and vibration transfer functions as well as source levels, e.g. shaking forces and noise radiation of powertrains. If any component significantly fails to meet its noise-related target, full vehicle noise performance will also probably not meet its targets. Typical virtual series deliverables for full vehicle noise and vibration refinement are:

•

Vehicle modal alignment: all component modes should properly be separated to avoid any 'resonance catastrophe' under operating conditions.

•

Vehicle sensitivity to 'unit excitation': this can be a vibration transfer function (vehicle response at the customer perception point due to, e.g., unit mount force, calculated by FE tools), a noise transfer function (p/F, also calculated by FE tools) or noise reduction (sound pressure difference between engine bay and passenger cabin, calculated by SEA tools).

•

Response to typical vehicle load cases, e.g. structure-borne response to idle excitation (loads of the powertrain applied to a full vehicle FE model), response to wheel imbalance (nibble, solved by FE or system dynamics tools depending on vehicle content for electronic tools to reduce steering wheel rotation – EPAS), response to road excitation or road noise (road surface irregularities applied to tyre patch points using an FE tool); vehicle acceleration noise (including all load paths in the TPA tool); and high-frequency noises like wind noise (derive the shape-related pressure fluctuation on the outer vehicle skin from CFD simulation and apply this as load for SEA analyses).

The 'status assessments' that need to be prepared for the virtual series report will be followed by further activities to identify the root causes for missing attribute targets. Transfer path analyses as well as modal contribution analyses are typical methods applied for these noise and vibration refinement investigations (see Plates VII and VIII between pages 114 and 115).

Based on such findings, high-level corrective actions like increasing local stiffness by adding reinforcements to the CAE model or shifting component modes in the respective CAE model will be investigated. If these changes solve the issues, these concepts will be communicated to the respective component areas to develop feasible solutions that need to be incorporated into the design for the next virtual series or for prototypes, as appropriate.

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B978184569497550008X

Verification and validation of flight control system airborne software

Yakui Gao , ... Chaoyou Zhi , in Test Techniques for Flight Control Systems of Large Transport Aircraft, 2021

3.3.3.1 Testing strategy

The best strategy of SCADE-based model testing is to have simulation testing of each functional node on the host or test bed as early as possible. In other words, once the SCADE model is established, the simulation verification of the model shall be started as soon as possible.

Based on previous model engineering experience, the following testing strategies are proposed:

1.

Static check is the first step of the whole SCADE testing, including checking the syntax and semantics (completed by SCADE editor), having walkthrough to check the rationality of the naming rules of variables and parameters in the model, having walkthrough to check if each functional node/module has detailed and specific description, having walkthrough to check if the model can trace to last layer of specification, and having walkthrough to check if the SCADE model structure and algorithm are reasonable.

2.

Check if the examples for simulation testing are designed based on requirements (detailed design documents of flight control system or control law design documents).

3.

If new functions or modifications are added to the model, regression testing shall be conducted for verification.

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B9780128229903000036

Contribution to Product Line Engineering

Jean-Luc Voirin , in Model-Based System and Architecture Engineering with the Arcadia Method, 2018

15.4.2 Building a base of reusable components

This approach aims to build components specifically with a view to their reuse in a large number of projects (it is often called "building for reuse").

The choice to reuse components can be opportunistic, benefiting from a particular project in which some components are identified as of more general interest. In this case, it involves extracting from the project model a description of the component but also of its environment, which makes reuse easy. This component's dependencies on its environment will be verified, especially by analyzing the model, which could lead to its content and perimeter being reviewed, the addition of elements from other neighbor components or its external interfaces; its non-functional properties should also be considered and capitalized on, such as its resource consumption, its criticality and certification levels, etc.

The choice to reuse components can also be deliberate and proactive, constructing an engineering dedicated to these re-usable components, with an adapted approach. In this case, it is generally desirable to include the constraints of the first user projects in the component's definition; it thus means conducting a "multicustomer" engineering (the user projects are customers for component engineering), which is similar to building a product line: comparing operational and functional needs, seeking to maximize shared elements, defining a behavior that satisfies these different needs and federating them. Either projects are similar enough to one another for the projects to have a shared solution, or it will be necessary to think of a product line approach applied to the component.

  ***

In both cases, it is recommended to build at least one model dedicated to the component's own engineering, and one model intended for its integration in a system model including this component in its architecture.

The component's EM – and more generally the associated engineering data – will include at least:

–

the operational and system needs to which this component responds;

–

the component's environment: the components with which it communicates, associated protocols and scenarios, hosting component if necessary;

–

the component's logical and physical architecture, including non-functional dimensions: required resources, performances, security or certification level, possible settings and parameterization, etc.

–

the body of requirements associated with the component, test campaigns and proofs, associated documentation, supporting justification files (reliability, security, etc.), possible user manual, etc. (see also Chapter 9).

The component's integration model is a sort of "caricature", a simplified version of it, with a less fine (coarser grain) level of detail, as it is intended to be inserted into the overarching system model, at a level just sufficient to make it possible to make decisions on the system engineering. This model can be restricted to the physical or logical architecture for example, but can also provide elements of operational or needs analysis, to which must be added other engineering artifacts useful at the system level (requirements, tests, possible FM of the component, etc.).

As far as possible, if a number of reusable components exist in a single context, it is useful also to have a global domain model, which is similar to that mentioned in a subtractive approach, and which will include an operational analysis and system needs analysis describing the main cases where components are used, and a logical and physical architecture describing the component's conditions and rules of assembly, to enable their global, rather than only unitary and separate reuse.

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B9781785481697500159

27th European Symposium on Computer Aided Process Engineering

Foteini Barla , ... Antonis Kokossis , in Computer Aided Chemical Engineering, 2017

2.2.1 Process Flowsheeting Modelling

Bespoke and commercial process engineering models are developed to combine the lab chemistries in RESYNTEX, process engineering evidence at pilot unit processes and new developments at the demonstration site. The flowsheeting models employs equilibrium data, reaction kinetics and mass transfer in order to establish mass and energy balances for the various hydrolysis, solubilisation and depolymerisation processes. Additional models address waste treatment processes, solvent extraction, and alternative separation processes that bear several technical and economic trade-offs. The modelling work incorporates important design parameters such as residence times, substrate purity/concentration, pH, viscosity etc., as they are required to configure later points of integration within the overall flowsheet. Model regression and parameter estimation is applied for maximising process efficiencies and minimising costs. All models are validated by the real-life pilot process.

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B978044463965350324X

Medical Biotechnology and Healthcare

J.I. Dawson , ... R.O.C. Oreffo , in Comprehensive Biotechnology (Second Edition), 2011

5.22.7 In vivo Models of Skeletal Regeneration

In vitro skeletal tissue engineering models have provided important information on the molecular/cellular interaction and mechanisms involved in osteogenesis. Recently, models have been developed with multicellular, three-dimensional components to aid our understanding of osteogenesis, yet, these in vitro models still cannot provide the complex physiological microenvironments for skeletal development and repair that can be provided by in vivo models. The vasculature plays a fundamental role in osteogenesis and this component, thus far, cannot be functionally reproduced in vitro. In vivo models of skeletal defects have allowed the investigation of a diverse assortment of engineered constructs to develop the most effective strategy for skeletal regeneration. These models provide the necessary blood supply, nutrients, gaseous exchange and waste removal, mechanical stimuli, multicellular/growth factor/tissue interactions all in a three-dimensional environment. When using in vivo models to explore the biocompatibility of bone tissue-engineered constructs, a number of parameters need to be considered such as breed of animal, age, sex, nutrition, size of animal, anesthetic, aftercare, mechanical loading, and fixation of the bone/construct. It is vitally important that the untreated bone defects (controls) in these models are of a critical size showing negligible repair and that the in vivo model procedures and analysis are robustly reproducible [19].

A number of in vivo models have been developed to investigate bone development and repair. The simplest model to evaluate bone engineering constructs in vivo is the subcutaneous implant model, where the investigative material, that is, tissue engineering composites are placed under the skin of an animal for a set period of time. This model allows for the simultaneous evaluation of angiogenesis and bone formation by perfused contrast dyes and microcomputed tomography, respectively. The disadvantages in employing this model are the infiltration of the host tissue and cells into the implanted material making it difficult to delineate the exact effect of the implanted material. Diffusion chamber models overcome this problem because of their enclosed environment, which allows the growth of transplanted cells, tissue, or biomaterials in a syngeneic or allogeneic animal host without growth and infiltration of the host cells into the implanted sample.

More complex in vivo models such as the calvarial defect model have been developed to evaluate bone tissue engineering strategies for skeletal repair. This robust model provides precise radiographic and histological evaluation of implanted tissue-engineered constructs within a standardized craniotomy defect, and because the dura and cranial skin act as support structures for the implant, bone support mechanisms and fixators are not required. The problem associated with this model is that unlike endochondral bone formation in long bones, cranial bone develops by an intramembranous mechanism and therefore is not subjected to load-bearing forces.

A number of load-bearing critical-sized femur defects in small animal models have been developed to assess the efficacy of bone-engineered strategies. The use of immunocompromized mouse models for nonunion critical-sized femur defects have allowed for the assessment of seeded human osteoprogenitors onto tissue-engineered composites with minimal inflammatory responses and rejection of the tissue-engineered constructs. Fixation of the bone in critical-sized defects, particularly in larger animal models, reduces any micromotion that the implanted composite may experience. Constructs for bone repair have been extrapolated to larger animal studies such as sheep to assess tissue-engineering construct implants in bone regeneration. For example, it was demonstrated that allograft implants can enhance bone repair in a 30-mm tibial segmental defect which had been fixated by interlocking intramedullary nails. Moreover, addition of other factors such as chondroitin sulfate enhances bone remodeling and bone formation around hydroxyapatite/collagen composites implanted into 3-cm critical-sized sheep defects. Combination strategies of osteoprogenitors and ceramic biomaterial constructs similar to those used in small animal critical-sized bone defect models have also been applied to sheep models. Interestingly, in these experimental models, increased vascular ingrowth was demonstrated in the implanted cell seeded scaffolds. Furthermore, transplanting blood-derived endothelial progenitor cells into a critical-sized sheep tibia defect can enhance the repair and regeneration.

Other in vivo models used to assess bone repair include the bone chamber model and DO. The bone chamber model in large animals has provided researchers valuable insights into the interactions of skeletal tissue, functional biomaterials, transplanted progenitor cells, and numerous growth factors. Furthermore, the advanced surgical technique of DO provides an ideal model for tissue regeneration and repair and has allowed researchers to gain a better understanding of the body's potential to self-renewal and self-repair providing insights for the development of new clinical skeletal tissue engineering strategies.

In vivo bone defect models without the use of fixators to evaluate the bone regenerative potential of implanted ceramic/PLA scaffolds and human fetal bone cells have also been developed. In rat models, drill hole defects in the cancellous bone of the femoral and distal condyles have been used to assess implanted tissue-engineered constructs. The advantage of these models is that soft tissue-engineered constructs can be placed in these skeletal defects that experience load-bearing forces without the construct being damaged by the forces of the bone tissue or the use of fixation apparatus. Soft-engineered scaffolds encapsulated with co-cultured cells can increase mineralization when implanted bilaterally into metaphyseal bone defect perforations (0.9   mm diameter). Furthermore, other investigators have been able to apply this drill defect model to assess biodegradable composites in larger animal models.

Due to the complex mineralized matrix composition of bone and its unique interaction with its surrounding tissues, novel in vitro and in vivo model systems are required for the development of new strategies for regenerative medicine. In vivo models need to evolve to satisfactorily address a number of tissue regenerative issues, particularly for the complex clinical bone regenerative strategies, where bone repair is functionally lacking. Such issues include the implantation and survival of stem cell/osteoprogenitor cells; growth factor release kinetics; biocompatibility and vascularization of engineered constructs; and standardized experimental design and analysis principally in larger animal models.

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B978008088504900430X

Process design for continuous separations

E.S. Tarleton , R.J. Wakeman , in Solid/Liquid Separation, 2007

Publisher Summary

This chapter describes the process engineering models for continuous filter cycles. These models facilitate detailed calculations and provide a platform for the development of computer simulations. The chapter describes the principal features of common continuous filter cycles and presents the equations required to model these cycles. It provides detailed example calculations for the horizontal belt filter and the rotary drum filter as these are representative of typical continuous cycles. The chapter shows how computer simulations can be used to examine in detail the effects of process variables on continuous filter performance. It provides additional information on the acceptable ranges for operational parameters in continuous filters, and outlines the principles involved in troubleshooting their operation.

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B9781856174213500079

4 Main Uses of Modeling in Engineering Design

Source: https://www.sciencedirect.com/topics/engineering/engineering-model